Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The naughty username checking system used by Twitch

ghostbin.com

Post reply on HN
Print view View original

291–300 of 348 posts

Re: The naughty username checking system used by Twitch

#291

Post by marcus_holmes »

pjc50 wrote: post #254
marcus_holmes wrote: post #225

Earlier quoted context omitted.

Is that just well-known in the UK though? I've mentioned this to people outside the UK and only occasionally have they known it.

Nobody outside the UK has heard of Scunthorpe because it can't make it through the filters. (There's a ridiculously large list of innuendo place names for tiny villages: https://anglotopia.net/ultimate-list-of-funny-british-place-... )

One of Douglas Adams' lesser-known but still hilarious works: https://www.panmacmillan.com/authors/douglas-adams/the-meani...

Re: The naughty username checking system used by Twitch

#292

Post by marcus_holmes »

chrisdengso wrote: post #274
marcus_holmes wrote: post #167

We used to call this the "Scunthorpe problem" - how do you filter for obviously rude names while still allowing people to have their actual names? Bearing in mind that some people's names are actually rude. I worked on the application form processing for the Nectar card launch in the UK back in the early 2000's, and we had several cases. Luckily we had human data-entry clerks in the loop, so all we had to do was flag…

Haha, I call this the "Clbuttic" problem. But nowadays it can be solved with machine learning fairly easy https://moderationapi.com/blog/moderate-text-automatically-u...

That's a slightly different problem from the Scunthorpe problem.

You have a customer who gives her name as "Fanny Batter". Is that a permissible name?

What about "Juan Kerr"? Or "Amanda Huggenkiss"?

It's not about replacing "ass" with "butt", but detecting rudeness/humour in context.

I'd be amazed if machine learning can solve this one. It's extremely hard for humans.

* all those names should be rejected btw

Re: The naughty username checking system used by Twitch

#293

Post by mnsc »

RHSeeger wrote: post #259
marcus_holmes wrote: post #237

Earlier quoted context omitted.

> But on a collective level we aren't yet (if we'll ever be) as developed a civilization to have empathy on that level. I agree. Which is why we have rules around politeness and manners, which roughly conform to same actions as actual empathy. Which makes sense - people don't always have the same level of empathy, there are lots of neurodivergent people who will never understand empathy but who can learn a set of rul…

> I feel like this respect for being nice to each other so we can all get along is on the wane. In my mind, part of that is that the rules are getting stricter, so people respect those rules less. Over time, the needle has moved further towards the idea that anyone can be offended by anything anyone else says. Even if it's clear (to the average person) that the speaker meant no harm (or even intended a totally differ…

> Over time, the needle has moved further towards the idea that anyone can be offended by anything anyone else says.

What do you mean by this? Are you talking about some contrived example where one person had their parents murdered by a maniac while shouting "bicycle" over and over which causes that word to be triggering ptsd for that individual. In that case that idea is correct since that, hypothetically, can happen for _any_ word yes. But how is that relevant to any of this?

Re: The naughty username checking system used by Twitch

#294

Post by ridaj »

OskarS wrote: post #162
ridaj wrote: post #79

Earlier quoted context omitted.

This looks like legit, no-nonsense gets-the-job-done code that gets updated every time some jerk find a new way to be a jerk to others. It isn't great, but at least not over-engineered, and I'm not sure if Twitch account sign-up volumes and abuse are at the point where they should staff a project to do this more robustly / scalably

Exactly what I think. This is the kind of code that doesn't have a platonic ideal, it has to get updated with time and experience and reports. There is no "non-hacky" way to do this, you just have to look at the reports that are coming in and keep adding rules that are relevant.

Well there are some significantly less hacky / more scalable ways to do this. This looks at the edge of maintainability, but if the lists were 10x as long and people constantly stepped on each other's toes causing outages while making updates, a bit of project investment probably wouldn't be a bad idea.

Re: The naughty username checking system used by Twitch

#296

Post by phamilton »

omegalulw wrote: post #117
phamilton wrote: post #34

We had to do this for a link shortening system (to make sure random base64 didn't contain profanity). It was a pretty fun problem. Not just the implementation, but doing the math to make sure it didn't make our shortened links easily enumerable. The implementation wasn't too bad, but we set up logging initially to spit out any random strings it decided to block. I demo'd this in front of the whole company and live ta…

How statistically likely is this? Can't you just regenerate until no you have a suitable short URL. Aside from performance, this is as random as you can be. Or generate the characters one by one and backtrack, this requires less random days. Or regenerate the unwanted substring.

Just trying again is absolutely fine.

Our main concern was whether we needed to increase the size to 26 to account for the loss of keys. After doing the math, a 25 digit random string has a ~5% chance of containing one of 150 three or four character inappropriate substrings. That 5% loss isn't that big of a deal. But we had to figure out the math as part of due diligence before shipping.

Re: The naughty username checking system used by Twitch

#297

Post by _hilro »

OJFord wrote: post #210
_hilro wrote: post #190

Earlier quoted context omitted.

The word is bitch so the sanitised version would be B** Why portray the bad word as 3 letters - it doesn't make sense.

Ha, the same thing probably happened to GP as has happened to you - unescaped * characters (escape with `\`) resulting in two imperceptibly italic asterisks.

you are 100% correct!

Re: The naughty username checking system used by Twitch

#298

Post by brokenwren »

We tried to convince Twitch for years that their filters were garbage and they should use CleanSpeak. They kept insisting their engineering team had written the best filter in the world.

Sometimes it’s just not worth the effort to try and help people solve these problems.

Re: The naughty username checking system used by Twitch

#300

Post by ridaj »

mc32 wrote: post #11
rhema wrote: post #5

I have a hard time believing this was / is the real version used. It doesn't seem broad enough. More likely it was a kind of smoketest that made sure that a more automated keyword checker was working. It does remind me of the XKEYSCORE (Snowden leaks) that used keywords to bubble up potential threats from emails etc https://www.businessinsider.com/nsa-prism-keywords-for-domes... .

It also mostly checks for English naughty words and not much else. People can have fun in lots of other languages, so it would seem this is a small sample.

Is Twitch's footprint big enough internationally that they have to worry about it?
Post reply on HN